The high sample complexity of reinforcement learning challenges its use in practice. A promising approach is to quickly adapt pre-trained policies to new environments. Existing methods for this policy adaptation problem typically rely on domain randomization and meta-learning, by sampling from some distribution of target environments during pre-training, and thus face difficulty on out-of-distribution target environments. We propose new model-based mechanisms that are able to make online adaptation in unseen target environments, by combining ideas from no-regret online learning and adaptive control. We prove that the approach learns policies in the target environment that can quickly recover trajectories from the source environment, and establish the rate of convergence in general settings. We demonstrate the benefits of our approach for policy adaptation in a diverse set of continuous control tasks, achieving the performance of state-of-the-art methods with much lower sample complexity.
Open-Set Domain Adaptation Under Background Distribution Shift: Challenges and A Provably Efficient Solution
Department of Computer Science, Johns Hopkins University（约翰霍普金斯大学计算机科学系）；Faculty of Data and Decision Sciences, Technion（技术学院数据与决策科学学院）；Center for Data Science, New York University（纽约大学数据科学中心）；Bayesian Health（贝叶斯健康）
University of Oxford（牛津大学）；Auburn University（奥本大学）；Huawei Technologies（华为技术有限公司）
We present the first provably convergent two-timescale off-policy actor-critic algorithm (COF-PAC) with function approximation. Key to COF-PAC is the introduction of a new critic, the emphasis critic, which is trained via Gradient Emphasis Learning (GEM), a novel combination of the key ideas of Gradient Temporal Difference Learning and Emphatic Temporal Difference Learning. With the help of the emphasis critic and the canonical value function critic, we show convergence for COF-PAC, where the critics are linear and the actor can be nonlinear.
CourtGuard: A Model-Agnostic Framework for Zero-Shot Policy Adaptation in LLM Safety
Department of Computer Science, Virginia Tech（弗吉尼亚理工大学计算机科学系）；School of IT（信息技术学院）；Engineering, ADA University（工程学院，ADA大学）；School of Law, ADA University（法学院，ADA大学）
Tsinghua University（清华大学）；Meta AI Research（Meta人工智能研究院）；Princeton University（普林斯顿大学）；University of Washington（华盛顿大学）
Policy-based methods with function approximation are widely used for solving two-player zero-sum games with large state and/or action spaces. However, it remains elusive how to obtain optimization and statistical guarantees for such algorithms. We present a new policy optimization algorithm with function approximation and prove that under standard regularity conditions on the Markov game and the function approximation class, our algorithm finds a near-optimal policy within a polynomial number of samples and iterations. To our knowledge, this is the first provably efficient policy optimization …
While deep reinforcement learning has successfully solved many challenging control tasks, its real-world applicability has been limited by the inability to ensure the safety of learned policies. We propose an approach to verifiable reinforcement learning by training decision tree policies, which can represent complex policies (since they are nonparametric), yet can be efficiently verified using existing techniques (since they are highly structured). The challenge is that decision tree policies are difficult to train. We propose VIPER, an algorithm that combines ideas from model compression and…
University of Texas at Austin（德克萨斯大学奥斯汀分校）；Tsinghua University（清华大学）；University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）
Off-policy evaluation provides an essential tool for evaluating the effects of different policies or treatments using only observed data. When applied to high-stakes scenarios such as medical diagnosis or financial decision-making, it is crucial to provide provably correct upper and lower bounds of the expected reward, not just a classical single point estimate, to the end-users, as executing a poor policy can be very costly. In this work, we propose a provably correct method for obtaining interval bounds for off-policy evaluation in a general continuous setting. The idea is to search for the …
We study decentralized policy learning in Markov games where we control a single agent to play with nonstationary and possibly adversarial opponents. Our goal is to develop a no-regret online learning algorithm that (i) takes actions based on the local information observed by the agent and (ii) is able to find the best policy in hindsight. For such a problem, the nonstationary state transitions due to the varying opponent pose a significant challenge. In light of a recent hardness result \citep{liu2022learning}, we focus on the setting where the opponent's previous policies are revealed to the…
We study the global convergence and global optimality of actor-critic, one of the most popular families of reinforcement learning algorithms. While most existing works on actor-critic employ bi-level or two-timescale updates, we focus on the more practical single-timescale setting, where the actor and critic are updated simultaneously. Specifically, in each iteration, the critic update is obtained by applying the Bellman evaluation operator only once while the actor is updated in the policy gradient direction computed using the critic. Moreover, we consider two function approximation settings …
Courant Institute of Mathematical Sciences, New York University（纽约大学库朗数学科学研究所）；Center for Data Science, New York University（纽约大学数据科学中心）；Institute for Advanced Study, Princeton（普林斯顿高等研究院）
Domain adaptation in imitation learning represents an essential step towards improving generalizability. However, even in the restricted setting of third-person imitation where transfer is between isomorphic Markov Decision Processes, there are no strong guarantees on the performance of transferred policies. We present problem-dependent, statistical learning guarantees for third-person imitation from observation in an offline setting, and a lower bound on performance in the online setting.
Policy optimization methods remain a powerful workhorse in empirical Reinforcement Learning (RL), with a focus on neural policies that can easily reason over complex and continuous state and/or action spaces. Theoretical understanding of strategic exploration in policy-based methods with non-linear function approximation, however, is largely missing. In this paper, we address this question by designing ENIAC, an actor-critic method that allows non-linear function approximation in the critic. We show that under certain assumptions, e.g., a bounded eluder dimension $d$ for the critic class, the …
Provably Efficient Off-Policy Adversarial Imitation Learning with Convergence Guarantees
Division of Systems Engineering（系统工程系）；Department of Computer Science（计算机科学系）；Amazon Robotics（亚马逊机器人）；Department of Electrical and Computer Engineering（电气与计算机工程系）
Google Brain Robotics（谷歌大脑机器人团队）；UC Berkeley（加州大学伯克利分校）；Columbia University（哥伦比亚大学）
Interest in derivative-free optimization (DFO) and "evolutionary strategies" (ES) has recently surged in the Reinforcement Learning (RL) community, with growing evidence that they can match state of the art methods for policy optimization problems in Robotics. However, it is well known that DFO methods suffer from prohibitively high sampling complexity. They can also be very sensitive to noisy rewards and stochastic dynamics. In this paper, we propose a new class of algorithms, called Robust Blackbox Optimization (RBO). Remarkably, even if up to $23\%$ of all the measurements are arbitrarily c…
Causality Meets Locality: Provably Generalizable and Scalable Policy Learning for Networked Systems
King’s College London（伦敦国王学院）；Eindhoven University of Technology（埃因霍温理工大学）；University of California, San Diego（加州大学圣地亚哥分校）；The Alan Turing Institute（艾伦·图灵研究所）
Risk-Sensitive Deep RL: Variance-Constrained Actor-Critic Provably Finds Globally Optimal Policy
Northwestern University（西北大学）；Peking University（北京大学）；University of Science and Technology of China（中国科学技术大学）；Duke University（杜克大学）；Yale University（耶鲁大学）；Pennsylvania State University（宾夕法尼亚州立大学）
We present Revel, a partially neural reinforcement learning (RL) framework for provably safe exploration in continuous state and action spaces. A key challenge for provably safe deep RL is that repeatedly verifying neural networks within a learning loop is computationally infeasible. We address this challenge using two policy classes: a general, neurosymbolic class with approximate gradients and a more restricted class of symbolic policies that allows efficient verification. Our learning algorithm is a mirror descent over policies: in each iteration, it safely lifts a symbolic policy into the …
University of California, Los Angeles（加州大学洛杉矶分校）；Microsoft Research（微软研究院）
We study lifelong reinforcement learning (RL) in a regret minimization setting of linear contextual Markov decision process (MDP), where the agent needs to learn a multi-task policy while solving a streaming sequence of tasks. We propose an algorithm, called UCB Lifelong Value Distillation (UCBlvd), that provably achieves sublinear regret for any sequence of tasks, which may be adaptively chosen based on the agent's past behaviors. Remarkably, our algorithm uses only sublinear number of planning calls, which means that the agent eventually learns a policy that is near optimal for multiple task…
University of Southern California（南加利福尼亚大学）；Facebook Inc.（脸书公司）；Princeton University（普林斯顿大学）；Northwestern University（西北大学）
We study the Safe Reinforcement Learning (SRL) problem using the Constrained Markov Decision Process (CMDP) formulation in which an agent aims to maximize the expected total reward subject to a safety constraint on the expected total value of a utility function. We focus on an episodic setting with the function approximation where the Markov transition kernels have a linear structure but do not impose any additional assumptions on the sampling model. Designing SRL algorithms with provable computational and statistical efficiency is particularly challenging under this setting because of the nee…
Ludwig-Maximilians University（路德维希-马克西米利安大学）；Technical University of Munich（慕尼黑工业大学）；Siemens Technology（西门子技术研究院）
Safe Policy Improvement (SPI) is an important technique for offline reinforcement learning in safety critical applications as it improves the behavior policy with a high probability. We classify various SPI approaches from the literature into two groups, based on how they utilize the uncertainty of state-action pairs. Focusing on the Soft-SPIBB (Safe Policy Improvement with Soft Baseline Bootstrapping) algorithms, we show that their claim of being provably safe does not hold. Based on this finding, we develop adaptations, the Adv-Soft-SPIBB algorithms, and show that they are provably safe. A h…
Temporal-difference learning (TD), coupled with neural networks, is among the most fundamental building blocks of deep reinforcement learning. However, due to the nonlinearity in value function approximation, such a coupling leads to nonconvexity and even divergence in optimization. As a result, the global convergence of neural TD remains unclear. In this paper, we prove for the first time that neural TD converges at a sublinear rate to the global optimum of the mean-squared projected Bellman error for policy evaluation. In particular, we show how such global convergence is enabled by the over…
University of Chicago（芝加哥大学）；Meta Platforms, Inc.（元平台公司）；University of Michigan（密歇根大学）；Northwestern University（西北大学）；Yale University（耶鲁大学）
While single-agent policy optimization in a fixed environment has attracted a lot of research attention recently in the reinforcement learning community, much less is known theoretically when there are multiple agents playing in a potentially competitive environment. We take steps forward by proposing and analyzing new fictitious play policy optimization algorithms for zero-sum Markov games with structured but unknown transitions. We consider two classes of transition structures: factored independent transition and single-controller transition. For both scenarios, we prove tight $\widetilde{\m…
Policy Optimization Provably Converges to Nash Equilibria in Zero-Sum Linear Quadratic Games
Department of Electrical and Computer Engineering & Coordinated Science Laboratory, University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校电子工程与协调科学实验室部门）；Department of Operations Research and Financial Engineering, Princeton University（普林斯顿大学运筹学与金融工程系）
The Hong Kong University of Science and Technology（香港科技大学）；Peking University（北京大学）；Fudan University（复旦大学）；Northwestern University（西北大学）；Yale University（耶鲁大学）
We study episodic two-player zero-sum Markov games (MGs) in the offline setting, where the goal is to find an approximate Nash equilibrium (NE) policy pair based on a dataset collected a priori. When the dataset does not have uniform coverage over all policy pairs, finding an approximate NE involves challenges in three aspects: (i) distributional shift between the behavior policy and the optimal policy, (ii) function approximation to handle large state space, and (iii) minimax optimization for equilibrium solving. We propose a pessimism-based algorithm, dubbed as pessimistic minimax value iter…
Batch reinforcement learning (RL) is important to apply RL algorithms to many high stakes tasks. Doing batch RL in a way that yields a reliable new policy in large domains is challenging: a new decision policy may visit states and actions outside the support of the batch data, and function approximation and optimization with limited samples can further increase the potential of learning policies with overly optimistic estimates of their future performance. Recent algorithms have shown promise but can still be overly optimistic in their expected outcomes. Theoretical work that provides strong g…
Model-free learning-based control methods have seen great success recently. However, such methods typically suffer from poor sample complexity and limited convergence guarantees. This is in sharp contrast to classical model-based control, which has a rich theory but typically requires strong modeling assumptions. In this paper, we combine the two approaches to achieve the best of both worlds. We consider a dynamical system with both linear and non-linear components and develop a novel approach to use the linear model to define a warm start for a model-free, policy gradient method. We show this…
We study Reinforcement Learning for partially observable dynamical systems using function approximation. We propose a new \textit{Partially Observable Bilinear Actor-Critic framework}, that is general enough to include models such as observable tabular Partially Observable Markov Decision Processes (POMDPs), observable Linear-Quadratic-Gaussian (LQG), Predictive State Representations (PSRs), as well as a newly introduced model Hilbert Space Embeddings of POMDPs and observable POMDPs with latent low-rank transition. Under this framework, we propose an actor-critic style algorithm that is capabl…
University of Maryland - College Park（马里兰大学帕克分校）；Cranberry-Lemon University（蔓越莓柠檬大学）；University of the Witwatersrand（威特沃特斯兰德大学）
The study of provable adversarial robustness for deep neural networks (DNNs) has mainly focused on static supervised learning tasks such as image classification. However, DNNs have been used extensively in real-world adaptive tasks such as reinforcement learning (RL), making such systems vulnerable to adversarial attacks as well. Prior works in provable robustness in RL seek to certify the behaviour of the victim policy at every time-step against a non-adaptive adversary using methods developed for the static setting. But in the real world, an RL adversary can infer the defense strategy used b…
We present an algorithm, HOMER, for exploration and reinforcement learning in rich observation environments that are summarizable by an unknown latent state space. The algorithm interleaves representation learning to identify a new notion of kinematic state abstraction with strategic exploration to reach new states using the learned abstraction. The algorithm provably explores the environment with sample complexity scaling polynomially in the number of latent states and the time horizon, and, crucially, with no dependence on the size of the observation space, which could be infinitely large. T…
Proximal Policy Optimization (PPO) is a highly popular model-free reinforcement learning (RL) approach. However, we observe that in a continuous action space, PPO can prematurely shrink the exploration variance, which leads to slow progress and may make the algorithm prone to getting stuck in local optima. Drawing inspiration from CMA-ES, a black-box evolutionary optimization method designed for robustness in similar situations, we propose PPO-CMA, a proximal policy optimization approach that adaptively expands the exploration variance to speed up progress. With only minor changes to PPO, our …
UC Berkeley（加州大学伯克利分校）；KIT（卡尔斯鲁厄理工学院）；Karlsruhe Inst. of Technology (KIT)（卡尔斯鲁厄理工学院）
Credit assignment in Meta-reinforcement learning (Meta-RL) is still poorly understood. Existing methods either neglect credit assignment to pre-adaptation behavior or implement it naively. This leads to poor sample-efficiency during meta-training as well as ineffective task identification strategies. This paper provides a theoretical analysis of credit assignment in gradient-based Meta-RL. Building on the gained insights we develop a novel meta-learning algorithm that overcomes both the issue of poor credit assignment and previous difficulties in estimating meta-policy gradients. By controllin…
Institute for Advanced Study（高等研究院）；Carnegie Mellon University（卡内基梅隆大学）；Princeton University（普林斯顿大学）；University of California, Los Angeles（加利福尼亚大学洛杉矶分校）
A fundamental challenge in artificial intelligence is to build an agent that generalizes and adapts to unseen environments. A common strategy is to build a decoder that takes the context of the unseen new environment as input and generates a policy accordingly. The current paper studies how to build a decoder for the fundamental continuous control task, linear quadratic regulator (LQR), which can model a wide range of real-world physical environments. We present a simple algorithm for this problem, which uses upper confidence bound (UCB) to refine the estimate of the decoder and balance the ex…
Robotics Innovation Center, DFKI GmbH（德国人工智能研究中心有限公司机器人创新中心）
Contextual policy search (CPS) is a class of multi-task reinforcement learning algorithms that is particularly useful for robotic applications. A recent state-of-the-art method is Contextual Covariance Matrix Adaptation Evolution Strategies (C-CMA-ES). It is based on the standard black-box optimization algorithm CMA-ES. There are two useful extensions of CMA-ES that we will transfer to C-CMA-ES and evaluate empirically: ACM-ES, which uses a comparison-based surrogate model, and aCMA-ES, which uses an active update of the covariance matrix. We will show that improvements with these methods can …
The ability to exploit prior experience to solve novel problems rapidly is a hallmark of biological learning systems and of great practical importance for artificial ones. In the meta reinforcement learning literature much recent work has focused on the problem of optimizing the learning process itself. In this paper we study a complementary approach which is conceptually simple, general, modular and built on top of recent improvements in off-policy learning. The framework is inspired by ideas from the probabilistic inference literature and combines robust off-policy learning with a behavior p…
We study offline reinforcement learning under a novel model called strategic MDP, which characterizes the strategic interactions between a principal and a sequence of myopic agents with private types. Due to the bilevel structure and private types, strategic MDP involves information asymmetry between the principal and the agents. We focus on the offline RL problem, where the goal is to learn the optimal policy of the principal concerning a target population of agents based on a pre-collected dataset that consists of historical interactions. The unobserved private types confound such a dataset …
We consider large-scale Markov decision processes (MDPs) with an unknown cost function and employ stochastic convex optimization tools to address the problem of imitation learning, which consists of learning a policy from a finite set of expert demonstrations. We adopt the apprenticeship learning formalism, which carries the assumption that the true cost function can be represented as a linear combination of some known features. Existing inverse reinforcement learning algorithms come with strong theoretical guarantees, but are computationally expensive because they use reinforcement learning o…
College of Intelligence and Computing, Tianjin University（天津大学智能与计算学部）；Northwestern Polytechnical University（西北工业大学）
Deep Reinforcement Learning (DRL) has been a promising solution to many complex decision-making problems. Nevertheless, the notorious weakness in generalization among environments prevent widespread application of DRL agents in real-world scenarios. Although advances have been made recently, most prior works assume sufficient online interaction on training environments, which can be costly in practical cases. To this end, we focus on an offline-training-online-adaptation setting, in which the agent first learns from offline experiences collected in environments with different dynamics and then…
University of Science and Technology of China（中国科学技术大学）；Northwestern University（西北大学）；Princeton University（普林斯顿大学）
In generative adversarial imitation learning (GAIL), the agent aims to learn a policy from an expert demonstration so that its performance cannot be discriminated from the expert policy on a certain predefined reward set. In this paper, we study GAIL in both online and offline settings with linear function approximation, where both the transition and reward function are linear in the feature maps. Besides the expert demonstration, in the online setting the agent can interact with the environment, while in the offline setting the agent only accesses an additional dataset collected by a prior. F…
This paper introduces Fast Linearized Adaptive Policy (FLAP), a new meta-reinforcement learning (meta-RL) method that is able to extrapolate well to out-of-distribution tasks without the need to reuse data from training, and adapt almost instantaneously with the need of only a few samples during testing. FLAP builds upon the idea of learning a shared linear representation of the policy so that when adapting to a new task, it suffices to predict a set of linear weights. A separate adapter network is trained simultaneously with the policy such that during adaptation, we can directly use the adap…
Learning Provably Stabilizing Neural Controllers for Discrete-Time Stochastic Systems
Sharif University of Technology（谢里夫理工大学）；Institute of Science and Technology Austria (ISTA)（奥地利科学技术研究所（ISTA））；Massachusetts Institute of Technology（麻省理工学院）
